ENSC 833: NETWORK PROTOCOLS AND PERFORMANCE

FINAL PROJECT PRESENTATION

SPRING 2016

EFFICIENT HANDOVER IMPLEMENTATION IN LTE-BASED FEMTO-CELL ENVIRONMENT

Presenters: Instructor :

Nnamdi Eyisi (neyisi@sfu.ca) Prof. Ljiljana Trakojovic

Haris Mehmood (hmehmood@sfu.ca)

Website:

http://www.sfu.ca/~hmehmood4/25/2016

1

OVERVIEW

Introduction

Project overview

LTE

Femto Cells

Types of handover

Implementation of proposed algorithm

Handover Algorithm

Handover Parameters

Results

Discussion

References

4/25/2016

2

INTRODUCTION

4/25/2016

3

PROJECT OVERVIEW

Proposed algorithm is to be implemented in an LTE based femtocell environment

Femtocells have low cost and power requirement, and relatively small coverage area. Typically

private owned and deployed indoors

Ping-pong effect usually more frequent femtocells largely due to its small coverage area

Handover of interest is User Equipment's (UEs) moving between femtocells and macrocells

Proposed algorithm aims to reduce frequent ping-pong handover using occurring at the boundary of

the cell

User Equipment's (UEs) that are candidates for handover are categorised first based on their speed

UEs within a certain speed range are given priority for handover depending on the type of service

4/25/2016

4

PING-PONG EFFECT

Ping-pong is common phenomenon which occurs as result

of unnecessary handovers and can reduce the system

efficiency by as far as 40%.

Our aim is to eliminate this problem in high speed

Vehicular/medium speed traffic environment.

4/25/2016

5

WHAT IS LTE ?

LTE (Long-Term Evolution , commonly marketed as 4G LTE) is a standard for

wireless communication of high-speed data for mobile phones and data terminals.

It is based on the GSM/EDGE and UMTS/HSPA network technologies, increasing the capacity and

speed using a different radio interface together with core network improvements.

The standard is developed by the 3GPP (3rd Generation Partnership Project) and is specified in its

Release 8 document series, with minor enhancements described in Release 9.

4/25/2016

6

FEMTO-CELL OVERVIEW

Femto-cells are defined as the low power cells that are normally used for indoor radio

reception.

They are normally part of macro-cells.

5G technologies are currently in the process of using millimeter wave and would deploy

femto-cells in commercial environment.

Our decision to study femto-cells stem from the fact that they part an important part of next

generation networks.

4/25/2016

7

FEMTO CELL ARCHITECTURE

Figure retrieved from www.telecom-cloud.net 4/25/2016

8

HANDOVER IN LTE

X2 Handover

This type of handover occurs between two macro cells belonging to the same Core network. X-

2 interface exists between the two e-nodeB.

S1 Handover

This type of handover occurs between macro cells belonging to the different core networks. It

flows through the core network instead of the radio network.

S1 handover is implemented between femtocells and macrocells as communication

between macrocells (eNB) and femtocells (HeNB) are routed via the core network and

not through the X2-interface (direct communication)

4/25/2016

9

S-1 BASED HANDOVER

Figure retrieved from jwcn.eurasipjournals.springeropen.com

4/25/2016

10

IMPLEMENTATION

4/25/2016

11

LENA HANDOVER ALGORITHMS

RSRP Algorithm

Reference Signal Received Power (RSRP), is defined as the linear average over the power contributions (in [W]) of the resource elements that carry cell-specific reference signals within the considered measurement frequency bandwidth.For RSRP determination the cell-specific reference signals R0 according TS 36.211 [3] shall be used. If the UE can reliably detect that R1 is available it may use R1 in addition to R0 to determine RSRP.

RSRQ Algorithm

Reference Signal Received Quality (RSRQ) is defined as the ratio N×RSRP/(E-UTRA carrier RSSI), where N is the number of RB’s of the E-UTRA carrier RSSI measurement bandwidth. The measurements in the numerator and denominator shall be made over the same set of resource blocks.

Noop Algorithm

No handover algorithm.

4/25/2016

12

Base LENA configuration

Soft Frequency Reuse

In Soft Frequency Reuse (SFR) scheme each eNb transmits over the entire system bandwidth, but there are two sub-bands, within UEs are served with different power level. Since cell-center UEs share the bandwidth with neighboring cells, they usually transmit at lower power level than the cell-edge UEs. SFR is more bandwidth efficient than Strict FR, because it uses entire system bandwidth, but it also results in more interference to both cell interior and edge users.

Proportional Fair Mac Scheduler

The Proportional Fair (PF) scheduling supports high resource utilization while maintaining good fairness among network flows. A user is likely to be scheduled when its instantaneous channel quality is high relative to its own average channel condition over time

4/25/2016

13

FEMTO CELL CREATION IN NS-3

Macro-cell

• Create a standard Macro cell

Assign Building Block

• Use building size equal to femto cell area

Place Femto Cell

Transmitter

• Isotropic antenna, Building loss model, Fading model, low power

4/25/2016

14

RADIO ENVIRONMENT MAP

The radio environment map shows three distinct macro cells, each having three sectors.

The box shows the building covered by the femto cells.

4/25/2016

15

CELL PARAMETER SELECTION

PARAMETER

NAME MACRO CELL FEMTO CELL

Cellular layout 14 three-sectored sites in hexagonal layout(42 cells in total) Single Cell (4 Placed inside each Macro cell)

Inter-site distance 500m 50m

Cell Tx power 30 dBm 10 dBm

Path loss model L = 128:1 + 37:6 log10 R Channel fading Typical urban Indoor propagation model with fading

Carrier frequency 2 GHz 2 GHz

System bandwidth 5 MHz (25 RBs) 5 MHz (25 RBs)

Error model None None

UE distribution 14 UEs distributed randomly in front of each eNodeB (588 UEs in total) 6 in each femto cell

UE measurement interval 25 ms 25 ms

Simulation duration 1200 s 1200 s

4/25/2016

16

MOBILITY PARAMETER SELECTION

PARAMETER

NAME VEHICULAR PEDESTRIAN URBAN

UE Type Outdoor Indoor Outdoor Indoor Outdoor Indoor

UE average speed 38 km/h 5 km/h 10 km/h 4 km/h 24 km/h 4 km/h

UE direction Coverage Edge Random Indoor Random Edge Random Indoor Random Random

Direction Change 5s 3s 5s 3s Random Random

UE type percentage 80% 20% 60% 40% 45% 55%

UE distribution Cluster Random Cluster Random Random Random

UE movement pattern Random Indoor path Random Indoor path Random Random

UE Min. Speed 25 km/h 2 km/h 2 km/h 2 km/h 15 km/h 0

UE Max. Speed 60 km/h 7 km/h 20 km/h 7 km/h 60 km/h 6 km/h

4/25/2016

17

TRAFFIC MODELING

PARAMETER

NAME UDP TCP TCP

Application Type udp echo server Voice Constant bit rate Bulk Send Application

Real Time No Yes No

Description Ping server VoIP Send Max Bytes

QCI No 3 No

Start time 0.0 0.0 0.0

End time 1200 1200 1200

Interval No No Burst traffic Reset 60s

Bandwidth Available EPC Guaranteed bit rate Maximum available bit rate

4/25/2016

18

PSEUDO CODEInitialization

Check for handover trigger conditions.

If(Handover threshold satisfied)

{Track the mobile speed using the Doppler shift

If(UE Speed is determined)

{ if !(Cell Max capacity)

{

If (UE Speed < 15km/h)

{Perform handover

{else if (UE Speed <=30km/h)

Determine whether the application is real time

If(Real Application)

{ Perform handover}

else {no handover}}}

If(UE Speed > 30km/h)

{ no handover }

}

else { no handover}}} 4/25/2016

19

4/25/2016

20

SIMULATION SCENARIOS

Scenario 1:

Apply the algorithm to the vehicular environment.

Scenario 2:

Determine the effect of different handover parameters in the pedestrian environment.

Scenario 2(a):

Analyze the effect of Time to trigger on the handover performance.

Scenario 2(b):

Analyze the effect of Threshold hysteresis on the number of handover.

Scenario 2(c):

Analyze the effect of window interval on the handover performance.

Scenario 3:

Apply the optimized parameters and algorithm to the urban environment and analyze the results.

4/25/2016

21

0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

0.9

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19

Handover Comparison

No of Handovers per user per second in 60s time intervals No of Handovers per user per second using efficient Algorithm in 60s time intervals

The graph shows that per user handover decreases considerably for high speed users using our

implemented algorithm. The performance improvement is almost 40 percent.

1.VEHICULAR ENVIRONMENT

a). Handover Comparison using algorithm

4/25/2016

22

-110

-105

-100

-95

-90

-85

-80

RSRP trace of a High speed user

Cell1 macrocell Cell2 femtocell

0

0.05

0.1

0.15

0.2

0.25

0.3

RA

DIO

LIN

K F

AIL

UR

E %

SIMULATION INTERVAL

Radio Link Failures Percentage

The radio failure percentage is limited to 0.25 percent which shows a lot of users are still able to

receive good radio reception of the macro cells when they are crossing the boundary.

The second graph shows the SINR trace of a single high speed user. This shows that the user though

has degraded experience for a few seconds is able to continue normal service. Since, there is no

handover the user is continually scanning for radio signals from both eNodeb and Home eNodeb.

b). Radio link Statistics

4/25/2016

23

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

60 120 180 240 300 360 420 480 540 600 660 720 780 840 900 960 1020 1080 1140

Handover

200 ms 400 msTime to Trigger

Ha

nd

ov

erp

er u

ser

2.PEDESTRIAN ENVIRONMENT

a). Time to Trigger

The graph shows that per user handover decreases considerably for pedestrian traffic when the time to trigger is increased, as the eNodeb takes more time to initiate handover procedure once the UE increases the offset threshold

4/25/2016

24

b). Handover Hysteresis

0

0.1

0.2

0.3

0.4

0.5

0.6

Time Interval

60

120

180

240

300

360

420

480

540600

660

720

780

840

900

960

1020

1080

Handover per user at Multiple Hysteresis Values

Hys=1 db Hys=3 db Hys=6 db

Increasing the threshold hysteresis decreases the overall number of handovers. However, this may

results in more radio failures and performance degradation.4/25/2016

25

c). Window Size

0

0.1

0.2

0.3

0.4

0.5

0.6

Sliding Window Size

Sliding Window size = 200ms 400ms

The graph shows that sliding window has a positive effect in decreasing the number of ping pong handovers. The sliding window is the time period where eNodeb averages the user power measurements to make a decision on handover.

4/25/2016

26

RADIO LINK FAILURES FOR WORST

CASE

0

0.1

0.2

0.3

0.4

0.5

0.6

Radio Link Failures Percentage

4/25/2016

27

3.URBAN ENVIRONMENT

a). Handover Parameter Calculation

Urban radio network is a combination of vehicular and pedestrian environment.

Determine the value of optimized handover parameters on the basis of the preceding environment

behavior.

Hysteresis = 3db. The optimized value for choosing the handover threshold.

Window size = a x W(ped)+b x W(veh)

whereas W(ped) corresponds to sliding window value for pedestrian environment. W(veh) corresponds to

the value of the parameter in the vehicular environment.

Trigger time = a x T(ped)+b x T(veh).

whereas T(ped) corresponds to handover trigger timer value for pedestrian environment. T(veh)

corresponds to the value of the parameter in the vehicular environment.

a and b are the fraction of UEs’ in the pedestrian and vehicular environment.

We will choose W=300 ms and T= 300ms.

4/25/2016

28

0

0.05

0.1

0.15

0.2

0.25

0.3

0.35

0.4

0.45

Handover vs Radio Link Failures

Handover per user per interval Radio Link Failures %

b). Handover Performance evaluation

The graph shows that the radio link failures do not exceed our benchmark of 0.5%.4/25/2016

29

CONCLUSION

We have finished the detailed modeling using vehicular and pedestrian environment and combining them both to form urban environment. It has given us insights on the radio behavior.

We demonstrated that proposed handover algorithm is efficient in reducing unnecessary handovers

The proposed algorithm had radio link failure rate of less than 0.3%

The categorization of network users based on speed and type of traffic reduced the amount of resources reduces the number of handovers while having minimal impact on quality of service

Every urban environment can be modeled as the combination of the vehicular and the pedestrian environment.

The optimization of the handover can be done by following the step by step approach.

High Speed vehicles passing through the boundary of femto-macro cell has still enough radio coverage to continue service reception.

4/25/2016

30

ALGORITHM PROPOSAL

Let Tv and Tp be the trigger time for the vehicular and pedestrian environment.

Let Wv and Wp be the sliding window averaging for the vehicular and pedestrian environment.

Alpha and Beta represent the fraction of UEs’ moving with less and more than 30 km/h respectively.

‘e’ represents the error during the previous computation.

Ts and Ws represent the stable values of trigger time and window averaging respectively.

Ti and Wi represent the indicated values of the above mentioned parameters.

Pp is defined as the function of ping pong handovers and pedestrian traffic. All handover between same

Ues’ within 2 seconds are determined as ping pong.

Pv is defined as the function of ping pong handovers and vehicular traffic.

For every averaging interval,

Determine Ti=alpha*Tp+Beta*Tv-e*Ti Wi=alpha*Wp+Beta*Wv-e*Wi

Error ‘e’= (Pp*alpha+Pv*beta)/Total Handovers e is positive if more errors are caused by vehicular

traffic and negative otherwise.4/25/2016

31

FUTURE RECOMMENDATIONS

Neural network based algorithm can be developed for self optimizing network.

Doppler shift can be implemented for calculating the velocity on the e-nodeb level.

Femto to femto handover can be further explored which has been our of the scope of this project.

User-user communication is also an emerging handover algorithm which can be efficient.

Application layer traffic can be changed to focus the studies more on the impact of type of traffic and

handover.

4/25/2016

32

REFERENCES

4/25/2016

33

[1] U. Dampage and C. Wavegedara, "A low-latency and energy efficient forward handover scheme for LTE-

femtocell networks", 2013 8th IEEE International Conference on Industrial and Information Systems,

Peradeniya, 2013, pp.53-58.

[2] C. Lee and J. Kim, "System Information Acquisition Schemes for Fast Scanning of Femtocells in 3GPP LTE

Networks", IEEE Commun. Lett., vol. 17, no. 1, pp. 131-134, Jan. 2013.

[3] A. Rath and S. Panwar, "Fast handover in cellular networks with femtocells", 2012 IEEE International

Conference on Communications (ICC), Ottawa, ON, 2012, pp.131-134.

[4] I. Shayea, M. Ismail and R. Nordin, "Advanced handover techniques in LTE- Advanced system", 2012

International Conference on Computer and Communication Engineering , Kuala Lumpur, 2012, pp.74-79.

[5] M. Lee and S. K. Oh, "Fast handover scheme using handover notification with no acknowledgement," 2011

IEEE MTT-S International Microwave Workshop Series on Intelligent Radio for Future Personal Terminals,

2011.

[6] Y. Wang, W. Peng, X. Qui, “Automated Handover Optimization Mechanism for LTE Femtocells” 7th

International ICST Conference on Communications and Networking in China (CHINACOM), 2012

[7] C. Lin, K. Sandrasegaran, H. Ramli, R. Basukala, “Optimized Performance Evaluation Of LTE Hard

Handover Algorithm With Average RSRP Constraint” International Journal of Wireless & Mobile Networks

(IJWMN) Vol. 3, No. 2, April 2011.

4/25/2016

34

[8] LTE Network Architecture. Retrieved, from http://www.tutorialspoint.com/lte/lte_network_architecture.htm

[9] H. Lee and Y. Lin, “A cache scheme for femtocell reselection,” IEEE Commununication Letter, vol. 14, no. 1, pp. 27–29, 2010.

[10] J. Moon and D. Cho, “Efficient handoff algorithm for inbound mobility in hierarchical macro/femto cell networks,” IEEE Communication Letter, vol. 13, no. 10, pp. 755–757, 2009.

[11] E. Dahlman, "LTE 3G Long Term Evolution," Expert Radio Access Technologies Ericsson Research, March 27, 2007.

[12] Z. Naizheng and J. Wigard„" On the Performance of Integrator Handover Algorithm in LTE Networks," in Proc. IEEE Vehicular Technology Conference. (VTC), pp. 1-5, 2008.

[13] I. Shayea et al,"Advanced handover techniques in LTE- Advanced system," in Proc. International Conference. on Computer and Communication. Engineering, pp.74-79, 2012.

[14] A. Rath et al., "Fast handover in cellular networks with femtocells," in Proc. 2012 IEEE International. Conference on Commununication. (ICC), pp. 2752-2757, June 2012.

[15] M.S. Shbat and V.Tuzlukov, "Handover technique between femtocells in LTE network using collaborative approach," in Proc. Asia-Pacific Conf. on Commun. (APCC), pp. 61-66, 2012.

[16] V. Chandrasekhar, "Femtocell networks: a survey," IEEE Commun.Mag., vol. 46, pp. 59-67, Sept 2008.

[17] C. Jiming, et al," Adaptive soft frequency reuse scheme for inbuilding dense femtocell networks," China Commun., vol. 10, no.1, pp. 44-55, 2013.

[18] K. Ivanov, G. Spring, ‘Mobile speed sensitive handover in a mixed cell environment’. Vehicular Technology Conference, July 1995, vol. 2, pp. 892–896

[19] W. Luo, X. Fang, M. Cheng, X. Zhou, ‘‘an optimized handover trigger scheme in LTE systems for high-speed railway’’. Signal Design and its Applications in Communications (IWSDA), October 2011, pp. 193–196

[20] P. Legg, G. Hui, J. Johansson, “A simulation study of LTE intra-frequency handover performance”. Vehicular Technology Conference Fall (VTC), September 2010, pp. 1–5.

[21] LENA ns-3 Retrieved, from http://www.nsnam.com

4/25/2016

35

QUESTIONS ?

4/25/2016

36